A consecutive series of 100 sleep apnea free patients with the complaint of excessive daytime somnolence (EDS) were evaluated; data from medical histories, physical examination, personality inventories, and polysomnography [nocturnal polysomnography (NPSG) and daytime multiple sleep latency testing (MSLT)] were tabulated. Significant differences were found between narcoleptic and non-narcoleptic patients in a number of parameters, including EDS severity, mean sleep latency on MSLT, sleep latency on NPSG, latency to REM sleep at night, number of REM sleep at night, number of REM sleep segments throughout the night, the total number of nocturnal myoclonic jerks (as well as the number occurring per hour of NREM and REM sleep), and the number of arousals and wake periods preceded by a myoclonic jerk. Significant differences in sleep latency during MSLT and NPSG testing were found between different EDS diagnostic groups of non-narcoleptic patients. The majority of patients in the MSLT group with long sleep latencies were in the diagnostic groups of EDS associated with psychophysiological and/or psychiatric problems or with drug abuse; patients with a diagnosis of idiopathic central nervous system hypersomnia or EDS associated with disturbed nocturnal sleep formed the majority of the MSLT group with short sleep latencies. The non-narcoleptic patients in a MSLT group with short sleep latencies had significantly shorter sleep latencies at night, more sleep cycles, higher sleep efficiency, and earlier REM sleep than patients with long sleep latencies.